WEBVTT Kind: captions; Language: en 00:00:03.000 --> 00:00:08.001 Hey, thanks for inviting me. Thanks everybody for coming. The re-decentralizing 00:00:08.001 --> 00:00:15.000 the web or the decentralized web is a topic here to my heart 00:00:15.000 --> 00:00:21.000 and it's great to have so much energy when Brewster starts anything, a lot of 00:00:21.000 --> 00:00:24.001 energy goes on. Just from having Brewster involved, never mind all the other 00:00:24.001 --> 00:00:28.000 people that he wrote in and we have a great crowd here today, so it's very 00:00:28.000 --> 00:00:34.000 exciting. So I suppose I'd been asked to as well start just a little bit, 00:00:34.001 --> 00:00:38.000 touch back on the original, you know what it was like back 00:00:38.000 --> 00:00:40.000 then, any other lessons. 00:00:40.000 --> 00:00:45.001 Were we supposed to be in your quads? No, you're not. But neither are you on my 00:00:45.001 --> 00:00:48.001 break. Maybe if you could just turn the projector off it would be 00:00:48.001 --> 00:00:50.001 good. Lots of house lights. 00:00:51.000 --> 00:00:56.000 This is a very yellow situation. It's also a really 00:00:56.000 --> 00:00:58.000 interesting quasi-religious sensation. 00:00:59.000 --> 00:01:03.000 You were saying that you worried about people who might think had weird dreams. 00:01:03.001 --> 00:01:10.000 Anybody who just woke up here in a yellow light filled church with sacred relics 00:01:10.000 --> 00:01:15.001 containing all of human knowledge at the back of the room flashing lights would 00:01:15.001 --> 00:01:22.000 definitely think they'd woken up in some kind of high-five rerun or something. 00:01:22.001 --> 00:01:29.001 This is very surreal, so very special to be here in this church of the church's 00:01:29.001 --> 00:01:36.000 real knowledge. So back then, okay, so Vint 00:01:36.000 --> 00:01:42.000 had done his work, thank you, Bob Kahn, folks, 20 years before I did the web 00:01:42.000 --> 00:01:46.001 thing. Okay, they did it. 69, okay, 89, totally different. 00:01:46.001 --> 00:01:51.000 Music, very different situations. It took 20 years for the people in Europe to 00:01:51.000 --> 00:01:54.000 actually get round to realizing that the Internet wasn't something, was something 00:01:54.000 --> 00:01:57.001 they could use even though they didn't invent it themselves. And so it became 00:01:57.001 --> 00:02:02.000 politically appropriate to use it as a concern. And the connectivity, which was 00:02:02.000 --> 00:02:07.000 more or less throughout the US, had actually percolated through often through 00:02:07.000 --> 00:02:14.000 people doing it kind of against the bosses' requests and actually running 00:02:14.000 --> 00:02:18.000 IP packets over lines which had been lent to them for other purposes, quietly 00:02:18.000 --> 00:02:22.001 tunneling and things. But it was there its own, and I could use it. And I had my 00:02:22.001 --> 00:02:27.001 boss, yeah, thought it was big, big, exciting, on that piece of paper which 00:02:27.001 --> 00:02:33.000 turned out only after he died, unfortunately, 10 years later, and he found that 00:02:33.000 --> 00:02:39.001 in his belongings. So at that point, the 00:02:39.001 --> 00:02:45.000 objective was to make something universal. I seen lots and lots of people make 00:02:45.000 --> 00:02:50.000 really, really cool documentation systems which required you to log into a 00:02:50.000 --> 00:02:55.001 particular system, use a particular format, use a particular server, use a 00:02:55.001 --> 00:03:00.001 particular client, use a particular word processor, or even use a particular 00:03:00.001 --> 00:03:04.000 structure, like some of them were tree-oriented, some of them are matrix 00:03:04.000 --> 00:03:09.001 -oriented. And the people, the guys who had matrix-oriented mind designed matrix 00:03:09.001 --> 00:03:12.001 -oriented documentation systems which you had to choose as sort of a 00:03:12.001 --> 00:03:16.001 classification and a category, and the people who are tree-oriented just rebelled 00:03:16.001 --> 00:03:21.000 and wouldn't use it. So it was this constant rebellion of people refusing to use 00:03:21.000 --> 00:03:26.000 documentation control systems and help systems. So the result was at CERN which 00:03:26.000 --> 00:03:31.000 had lots of really cool people, there was this mass of heterogeneity. And yeah, 00:03:31.001 --> 00:03:35.001 just stepping back from it, realizing actually all these systems different though 00:03:35.001 --> 00:03:40.000 they were, they all could be viewed as just hypertext pages. And so in fact, it 00:03:40.000 --> 00:03:44.001 would just take a bit of code to put on, to run on those systems, to convert them 00:03:44.001 --> 00:03:51.000 into hypertext pages within this global system. Those, they are of getting that 00:03:51.000 --> 00:03:57.000 actually to take off, which tricky, involved making a lot of, some, 00:03:58.000 --> 00:04:05.000 to a certain extent, a lot of compromises, like the data format HTML that looked 00:04:05.000 --> 00:04:10.000 like HTML, particular HTML, the way CERN people were using it, the tag h1 was 00:04:10.000 --> 00:04:14.001 because the people I worked with were used to writing h1 in front of a heading, 00:04:15.000 --> 00:04:18.001 and so they go along with that. They look at it and say, yeah, okay, that's, you 00:04:18.001 --> 00:04:20.001 know, HTML, whatever, I can do that. 00:04:21.000 --> 00:04:25.000 And the HTTP language, you know, so that is using angle brackets, so why 00:04:25.000 --> 00:04:29.000 shouldn't the network protocol use angle brackets too? Well, that's because the 00:04:29.000 --> 00:04:35.000 network protocols didn't. But if you look to SNTP and NNT, the network news, all 00:04:35.000 --> 00:04:40.001 existing protocols, ones that carried mail and carried the Usenet News, those, 00:04:41.000 --> 00:04:49.000 old enough to remember, Usenet News back there, great system, those protocols all 00:04:49.000 --> 00:04:53.000 had a different way of being, they had a sort of, you know, an uppercase verb, 00:04:54.001 --> 00:05:00.001 and the lines, and they sent, for example, things like mail messages around. So 00:05:00.001 --> 00:05:08.000 HTTP was just a concoction of SMTP and NNTP with a bit, and so when internet 00:05:08.000 --> 00:05:11.001 programmers back then looked at it, they looked at it, they go, oh, no, I can do 00:05:11.001 --> 00:05:17.000 that, yeah, that's straightforward, you're not trying to pull me into some weird 00:05:17.000 --> 00:05:22.001 space. So all, and the URL, first thing I had to have that we call URLs now, but 00:05:22.001 --> 00:05:31.000 they were using universal document identifiers back then, and the UDI was 00:05:31.000 --> 00:05:34.001 designed to look as much as possible like a Unix file name. I must admit I had 00:05:34.001 --> 00:05:41.000 no, no compunction at all about making the sashes go the right original way, even 00:05:41.000 --> 00:05:45.001 though a lot of the world at that point were putting the man backwards for 00:05:45.001 --> 00:05:50.001 reasons that historians will just write books about, I imagine. But so the 00:05:50.001 --> 00:05:54.001 sashes, it was great to have the sort, the sashes going the right way, but then, 00:05:55.000 --> 00:06:01.000 final names were good up to the point where you got to the root of your computer, 00:06:01.001 --> 00:06:06.000 and then there was a question of what to do about then the network piece of it, 00:06:06.000 --> 00:06:10.000 and the double slash actually came from the Apollo domain system. The Apollo 00:06:10.000 --> 00:06:15.001 domain, folks on a local area network, you could address files on other systems 00:06:15.001 --> 00:06:19.001 by putting a double slash computer name, so the idea was, again, anybody who'd 00:06:19.001 --> 00:06:30.000 seen the Apollo domain system would use that. So a lot of the web design was 00:06:30.000 --> 00:06:34.001 about picking up things that people had already used and putting them together so 00:06:34.001 --> 00:06:39.000 that people wouldn't feel that it was strange. I just mentioned that as one of 00:06:39.000 --> 00:06:44.001 the things which is useful, and if we can then build new systems, put new systems 00:06:44.001 --> 00:06:48.000 together, we'll have to probably use that sort of technique. Incremental change, 00:06:48.001 --> 00:06:54.001 only change the pieces that we need. But then, what happened was the thing that 00:06:54.001 --> 00:07:00.000 actually did take off. It was never clear that it was going to take off. Now 00:07:00.000 --> 00:07:02.001 people are really human, it's a done deal, but still you have to be careful the 00:07:02.001 --> 00:07:10.000 ways it can still fail. But when it took off, so then to rewind you back to the 00:07:10.000 --> 00:07:21.000 first, you know, first thing you 00:07:21.000 --> 00:07:28.001 [...] off. So you have to take off. So you have to take off. 00:07:28.001 --> 00:07:30.000 So you have to take 00:07:30.000 --> 00:07:33.000 off. So you have to take off. So you have to take off. So you have to take off. 00:07:33.000 --> 00:07:35.001 So you have to take off. So you have to take off. So you have to take 00:07:35.001 --> 00:07:38.001 off. So you have to take off. 00:07:40.000 --> 00:07:41.001 So you have to take off. So you have to take off. So you have to take off. So you 00:07:41.001 --> 00:07:41.001 have to take off. 00:07:41.001 --> 00:07:46.001 So you have to take off. So you have to take off. So you 00:07:46.001 --> 00:07:48.001 have to take off. So you have to take off. 00:08:13.000 --> 00:08:18.000 [...] off. So you have to take off. So you have to take off. So you have to take 00:08:19.000 --> 00:08:43.001 [...] to take off. So you have to take off. So you have to take off. So you have 00:08:43.001 --> 00:08:48.000 have to take off. So you have to take off. 00:08:48.000 --> 00:09:15.001 [...] 00:09:18.000 --> 00:09:21.000 The great thing about when inventing the platform, it didn't think about 00:09:21.000 --> 00:09:22.001 good thing about platform doesn't have attitude. 00:09:22.001 --> 00:09:24.001 It didn't have any attitude about what you should use it for. 00:09:25.000 --> 00:09:26.000 Similarly, the web. 00:09:26.001 --> 00:09:28.001 Now, no attitude about what you should use it for, and 00:09:28.001 --> 00:09:30.000 therefore no constraints on what you use it for. 00:09:30.001 --> 00:09:31.000 It's universal. 00:09:31.001 --> 00:09:33.001 You can do whatever you like, and therefore massive amount 00:09:33.001 --> 00:09:35.001 of creativity and innovation. 00:09:36.000 --> 00:09:39.001 And the creativity and innovation now has produced wonderful things like very 00:09:39.001 --> 00:09:43.000 powerful social networks where people spent a lot of their time. 00:09:43.000 --> 00:09:49.001 And until recently, but recently, for the last few years, last five years, last 00:09:49.001 --> 00:09:56.000 two years particularly, people have been coming up to me, and I've realized that 00:09:56.000 --> 00:09:58.001 out there, there's a sort of unease. 00:09:59.001 --> 00:10:05.000 There's an unease because there's people thinking, wait a moment, that utopian 00:10:05.000 --> 00:10:11.001 leveling of society and that reinvention of our systems of governance and our 00:10:11.001 --> 00:10:13.000 systems of debate. 00:10:14.001 --> 00:10:15.000 What happened to that? 00:10:16.000 --> 00:10:20.000 And the answer is, well, the people that I thought, we hoped we'd be making their 00:10:20.000 --> 00:10:22.001 own websites, and we imagine, fancily, putting them on the 00:10:22.001 --> 00:10:24.000 computer sitting under the TV. 00:10:25.000 --> 00:10:27.001 Actually, no, actually, they're all just on one great big website. 00:10:28.001 --> 00:10:32.000 And it turns out that when you talk to them, they're quite frustrated because you 00:10:32.000 --> 00:10:37.001 find a typical day in the life of a person out there. 00:10:37.001 --> 00:10:43.001 They have their friends on Facebook, and they have some photos on Flickr, and 00:10:43.001 --> 00:10:50.001 they have their colleagues are all on LinkedIn, and then all they want to do is 00:10:50.001 --> 00:10:55.000 they've uploaded some photos to Flickr, and they just want to share those photos 00:10:55.000 --> 00:10:57.001 with their colleagues on LinkedIn and with their 00:10:57.001 --> 00:11:00.000 friends who happened to be on Facebook. 00:11:01.000 --> 00:11:01.001 And you know what? 00:11:02.000 --> 00:11:02.001 Take out. 00:11:04.000 --> 00:11:07.000 And that's really, when you look at it, it's really stupid. 00:11:07.000 --> 00:11:12.000 But if you're in this world, of course, no, you have to either tell Flickr about 00:11:12.000 --> 00:11:16.000 all your friends and all your colleagues, or you have to move the photos onto 00:11:16.000 --> 00:11:21.001 Facebook, or you have to build a... 00:11:21.001 --> 00:11:22.001 You have to move your photos onto LinkedIn 00:11:22.001 --> 00:11:24.001 separately and send them out separately. 00:11:25.000 --> 00:11:29.001 Or you can build a third application and get... 00:11:29.001 --> 00:11:34.001 And run it and get it the right to extract the list of friends from Facebook, and 00:11:34.001 --> 00:11:38.000 you can try to build a third application which will build a bridge 00:11:38.000 --> 00:11:40.000 between those two, but then you have to learn... 00:11:40.000 --> 00:11:42.000 If you're a developer, you have to learn the Facebook API, and you have to learn 00:11:42.000 --> 00:11:44.000 the LinkedIn API, and you have to learn the Flickr API. 00:11:45.000 --> 00:11:46.000 And you know what? 00:11:46.000 --> 00:11:46.001 You know what? 00:11:46.001 --> 00:11:53.001 I'll just mail them to them with a stamp on it, because this, all my 00:11:53.001 --> 00:11:56.000 life, has been stuck in these silos. 00:11:57.000 --> 00:12:01.000 So the siloization of the web is what people are suddenly 00:12:02.000 --> 00:12:03.001 finding...are starting to react against. 00:12:03.001 --> 00:12:04.001 It was a great... 00:12:04.001 --> 00:12:09.001 If I had one side, it would be the cartoon from the Economist did several years 00:12:09.001 --> 00:12:11.001 ago now about people 00:12:14.001 --> 00:12:19.000 showing how, like prison cells, the world has got these walls. 00:12:19.001 --> 00:12:23.000 When you're in one social network, it was really hard to get into the other one, 00:12:23.000 --> 00:12:27.001 of people walking on top of the walls trying to help friends up and over into the 00:12:27.001 --> 00:12:29.000 next social network. 00:12:29.001 --> 00:12:30.000 It's like a great cartoon. 00:12:30.000 --> 00:12:34.001 [...] 00:12:34.001 --> 00:12:37.001 Here we are at a situation where the internet was 00:12:37.001 --> 00:12:39.001 designed to be basically decentralised. 00:12:39.001 --> 00:12:41.001 We can talk about that in detail. 00:12:42.000 --> 00:12:45.000 And Vince gone into some of the issues with DNS. 00:12:45.001 --> 00:12:52.000 And the web was designed to be decentralised, so everybody could participate by 00:12:52.000 --> 00:12:55.000 having their own domain, and having their own web server, and this wasn't how. 00:12:55.000 --> 00:13:02.000 And in fact, we've got a situation where individual personal data... ..has been 00:13:02.000 --> 00:13:03.001 locked up in these silos. 00:13:04.001 --> 00:13:06.000 And there is... 00:13:06.000 --> 00:13:09.001 And one of the features of the last few years is that if you read the trade 00:13:09.001 --> 00:13:14.001 press, you read magazines, what... 00:13:14.001 --> 00:13:18.001 The first thing you get on top of a lot of the articles, and certainly anything 00:13:18.001 --> 00:13:25.000 in the marketing world, is you get the assumption, they say, well, what the 00:13:25.000 --> 00:13:31.001 world today is one where the consumer gets a whole lot of things for free, 00:13:32.000 --> 00:13:39.000 and the deal that consumer has done is that they sold their soul of their privacy 00:13:39.001 --> 00:13:41.001 to in order to get those things for free. 00:13:42.000 --> 00:13:49.000 And when they sell it, when they give away the privacy to the machine, 00:13:49.001 --> 00:13:53.001 to the marketing system, to whoever it is that eventually gets the data that has 00:13:53.001 --> 00:13:57.000 been captured by you in all these many ways. 00:13:58.000 --> 00:14:02.001 And what happens to that allows them to suddenly get targeted with all kinds of 00:14:02.001 --> 00:14:08.001 things, becomes impossible to not let that person not let you... 00:14:08.001 --> 00:14:11.000 Not know that you're pregnant, so you end up getting... 00:14:11.000 --> 00:14:17.000 And your child ends up getting things for the rest of their life from just when 00:14:17.000 --> 00:14:23.000 your buying patterns switch, from going out a lot needing to buying diapers, or 00:14:23.000 --> 00:14:27.000 buying... You buy the first crib, because the system actually makes a huge amount 00:14:27.000 --> 00:14:30.001 of use of that data which people find suddenly a little bit creepy. 00:14:31.000 --> 00:14:34.001 But the assumption is that, but yeah, everybody's happy with that. 00:14:35.000 --> 00:14:35.001 That is the deal. 00:14:36.001 --> 00:14:40.000 They're not totally happy with that, but they're so happy to get a search for 00:14:40.000 --> 00:14:43.001 free and use for free that that's... 00:14:43.001 --> 00:14:46.001 And that's how it works. 00:14:46.001 --> 00:14:52.000 And the only way to make advertising on the net at the moment is by... To make 00:14:52.000 --> 00:14:53.001 money is to make advertising. 00:14:54.001 --> 00:14:57.001 So what's wrong with this picture? 00:14:58.000 --> 00:15:01.001 What's wrong with this picture is that... 00:15:01.001 --> 00:15:06.000 Well, I think start with that whole deal, that the deal that consumer has gone 00:15:06.000 --> 00:15:08.001 through, I put it to you, it's a myth. 00:15:09.000 --> 00:15:11.001 It's a myth that it has to be. 00:15:12.000 --> 00:15:14.000 I think it's a myth that everybody's happy with it. 00:15:14.000 --> 00:15:16.000 And I think it's a myth that it's optimal. 00:15:17.000 --> 00:15:23.001 So, in fact, what's weird about it is if you think of me as a consumer, right, my 00:15:23.001 --> 00:15:25.000 data... I go for a run. 00:15:25.001 --> 00:15:30.001 I go for a run, say, using some gadget, and my gadget communicates by hand to my 00:15:30.001 --> 00:15:35.000 Bluetooth, when my laptop, my laptop uploads it to the cloud, and then that cloud 00:15:35.000 --> 00:15:38.001 knows all about how fast I ran today. 00:15:39.000 --> 00:15:39.001 Well, you know what? 00:15:40.001 --> 00:15:45.001 The value of that little piece of data about how well I ran today, is about 00:15:45.001 --> 00:15:47.000 some guy with some demographic. 00:15:48.000 --> 00:15:52.001 And it's a little bit about how some guy in some particular demographic is 00:15:52.001 --> 00:15:56.000 choosing to keep fit today, for the point of view of marketers. 00:15:56.001 --> 00:16:03.000 It tells you a little bit about what I might buy, what I might sell me. 00:16:03.001 --> 00:16:04.001 Want to sell me some running shoes? 00:16:04.001 --> 00:16:06.000 Because you can figure out... 00:16:06.000 --> 00:16:09.000 Maybe you can figure out, by looking at the speed I'm running, that I probably 00:16:09.000 --> 00:16:12.000 need some new shoes, because I'm getting slower, because I'm limping or 00:16:12.000 --> 00:16:14.000 something, by looking at the GPS traces. 00:16:14.000 --> 00:16:20.000 But basically, the actual value to you, the big marketing machine, is actually 00:16:20.000 --> 00:16:24.001 really, really small, compared that actually, when you look at my point of view, 00:16:27.000 --> 00:16:29.000 that's data about me, OK? 00:16:30.000 --> 00:16:31.000 So if I'm... 00:16:31.000 --> 00:16:34.000 Supposing I'm running a machine trying to track up my life, that data... I put it 00:16:34.000 --> 00:16:38.000 to you, the value of the data to me is much greater than 00:16:38.000 --> 00:16:39.001 the value of the data to anybody else. 00:16:46.001 --> 00:16:47.000 So... not for a bit. 00:16:48.000 --> 00:16:49.000 We won't go for questions for a bit. 00:16:50.000 --> 00:16:51.001 Yeah, I know we have to catch up sometime, but we didn't talk 00:16:51.001 --> 00:16:53.000 about selling times at ending times. 00:16:57.001 --> 00:16:58.000 So... 00:16:58.000 --> 00:17:01.000 That would give us a half-nout of questions, if we went to questions now. 00:17:03.000 --> 00:17:04.000 So the... 00:17:06.000 --> 00:17:06.001 So... 00:17:08.001 --> 00:17:13.001 So the proposal is then to... bring back the idea of a decentralized 00:17:13.001 --> 00:17:15.001 web, bring back power to people. 00:17:16.000 --> 00:17:22.001 The things we're doing at MIT, various other organisations, involve 00:17:22.001 --> 00:17:27.000 just breaking that model, where all your data is in different 00:17:27.000 --> 00:17:29.001 silos, it involves making... 00:17:29.001 --> 00:17:31.001 We have a project, for example, which is called SOLID, 00:17:35.000 --> 00:17:40.001 and it... which involves having a standard, so that when I get my 00:17:40.001 --> 00:17:46.001 fit... data, whatever it is, I can put it in my own storage, I can put it in 00:17:46.001 --> 00:17:50.001 Cloud Storage, to iControl, and so we are thinking, we're going to make a social 00:17:50.001 --> 00:17:54.000 revolution by just tweaking, we're going to use web technology, but we're going 00:17:54.000 --> 00:17:58.000 to use it in a way that we can separate the apps that you use from the data that 00:17:58.000 --> 00:18:00.001 you use, and so that I can control my data. 00:18:00.001 --> 00:18:02.000 Imagine I have a place to keep my data, 00:18:05.000 --> 00:18:12.000 and... so the... whenever I run any app, it will store data 00:18:12.000 --> 00:18:16.000 in places I control, and as a result, when I run something like a Lifestyle app, 00:18:16.000 --> 00:18:22.001 it doesn't just get access to the things in one silo, like my friends, my 00:18:22.001 --> 00:18:27.000 management app, doesn't just get access to the stuff in my Facebook silo, it gets 00:18:27.000 --> 00:18:32.000 access to my whole life, and it gets access to the pieces of other people's lives 00:18:32.000 --> 00:18:34.001 that I'm happy to share with. 00:18:34.001 --> 00:18:38.001 So that's the way in which we're proposing, tweaking the architecture a little 00:18:38.001 --> 00:18:41.001 bit, but dramatically changing the way it's used socially. 00:18:42.000 --> 00:18:45.001 There are lots of other ways in which... we've had the Decentralized Interest 00:18:45.001 --> 00:18:51.001 Information Group at MIT for years, I've been talking about re-decentralizing the 00:18:51.001 --> 00:18:56.001 web, and there's a re-decentralized. org Move group, which has had some meetings 00:18:56.001 --> 00:19:02.000 in London, this... but when you use the word decentralized, you put up a 00:19:02.000 --> 00:19:05.001 workshop, then you meet all kinds of interests, and people who are looking at 00:19:05.001 --> 00:19:10.000 issues, looking at some of the issues, for example, with the existing web design, 00:19:10.001 --> 00:19:16.001 the fact that we use domain names system, well, you know, it was there, and it 00:19:16.001 --> 00:19:20.001 was clear that it has issues, as Vince mentioned. 00:19:21.001 --> 00:19:24.001 One of the things we could do is we could change the rules for domain name 00:19:24.001 --> 00:19:30.000 systems, so we... imagine we have a . archive in top-level domain where, actually, 00:19:30.001 --> 00:19:35.000 you don't rent them, you get them forever, and the deal is, if you have anything 00:19:35.000 --> 00:19:40.001 that's put in the .archive domain, then it's out there to be copied, it's subject 00:19:40.001 --> 00:19:46.000 to these interesting protocols, and that domain stays forever. 00:19:46.001 --> 00:19:50.001 You don't might... you might not keep it, you may die, but the domain stays and 00:19:50.001 --> 00:19:55.000 it will be managed by somebody else, and if nobody's using it actively, then 00:19:55.000 --> 00:20:00.000 it'll turn... it'll just go into the system of all the connected Internet 00:20:00.000 --> 00:20:01.001 archives in the world. 00:20:02.001 --> 00:20:08.000 So there are lots of things... there are lots of also really cool systems where 00:20:08.000 --> 00:20:13.001 you say, let's not use human-readable names which have semantics at all, because 00:20:13.001 --> 00:20:16.001 there's a kind of risky... people... 00:20:16.001 --> 00:20:22.000 By the way, people... you should think of the HTTP URL as a 00:20:22.000 --> 00:20:24.000 name, not as a location. 00:20:26.001 --> 00:20:28.001 Philosophically, it is a name, it is not a location. 00:20:29.000 --> 00:20:33.001 The fact is it's called a URL, it's because of a battle I lost in the IETF many 00:20:33.001 --> 00:20:39.000 years ago, but you should think about it as a name, and one of the ways to get 00:20:39.000 --> 00:20:43.000 people out of the thinking that these things are locations which can change is to 00:20:43.000 --> 00:20:47.001 actually not use the domain name at all, but to just use, say, a cryptographic 00:20:47.001 --> 00:20:54.001 hash of the actual text of the thing, 00:20:55.000 --> 00:20:58.001 which has lots of interesting properties and other things how you can generate a 00:20:58.001 --> 00:21:02.001 public private key pair for it and have it identified by the public key when you 00:21:02.001 --> 00:21:06.000 keep the private key, and that allows you to change it, and you can put the 00:21:06.000 --> 00:21:10.000 changes into the blockchain if you believe that the blockchain isn't another 00:21:10.000 --> 00:21:12.001 centralization that we're liable to fall into, 00:21:13.000 --> 00:21:14.001 if there's only just one blockchain. 00:21:15.000 --> 00:21:18.000 So lots and lots and lots around here, there are lots and lots and lots of it. 00:21:18.001 --> 00:21:24.001 We had some discussion at a pre-meeting yesterday with some very, very cool 00:21:24.001 --> 00:21:31.001 projects, which I'm sure they'll be available, but 00:21:31.001 --> 00:21:38.001 I think what we're all aiming for is to make the web 00:21:38.001 --> 00:21:44.000 more better in lots of ways, more reliable, more... 00:21:44.000 --> 00:21:45.001 [...] 00:21:45.001 --> 00:21:47.000 Give the ability when you follow a link. 00:21:47.000 --> 00:21:51.000 Already, for example, HTML, you can now... 00:21:51.000 --> 00:21:53.000 One little thing that's happened on the site, you can put 00:21:53.000 --> 00:21:54.001 a sub-resource integrity. 00:21:57.000 --> 00:22:01.000 You can just mention, just saying, that when you are linking to something with 00:22:01.000 --> 00:22:03.000 HTTP or R, you can actually give it a hash as well. 00:22:04.000 --> 00:22:06.000 So that's creeping into... 00:22:06.000 --> 00:22:08.000 You can talk to Wendy when it's down there. 00:22:08.001 --> 00:22:13.000 That's some new standards out from WTC, which are adding to the web as we know 00:22:13.000 --> 00:22:19.000 it, the sort of the address by hash, if you like, 00:22:19.001 --> 00:22:24.000 properties that a lot of you have got in these Wike Cool Decentralized systems. 00:22:25.000 --> 00:22:30.001 So a lot of the issues which we've been talking about, we will be talking about 00:22:30.001 --> 00:22:34.000 naming, about should we fix DNS. 00:22:35.001 --> 00:22:38.001 I think a lot of people feel that the squatting... 00:22:38.001 --> 00:22:39.001 I know when I've tried to... 00:22:39.001 --> 00:22:44.000 I had a really cool project and I've had this name for it, like When Can We Meet? 00:22:44.000 --> 00:22:48.000 I get WhenCanWeMeet.com and I find everything like WhenCanWeMeet.com has been 00:22:48.000 --> 00:22:51.001 bought by a speculator and has sent it to me for $100,000. 00:22:52.000 --> 00:22:54.000 That is not very good for innovation. 00:22:54.001 --> 00:22:59.000 If there is $100,000 tax for having a good idea, which consists of several 00:22:59.000 --> 00:23:01.001 English words, so obviously you say, When Can We... 00:23:01.001 --> 00:23:02.000 [...] 00:23:03.000 --> 00:23:07.000 And that's why the world is full of When Can We Mutt and things out there, and 00:23:07.000 --> 00:23:09.001 it's largely because of speculation, which kind of sucks. 00:23:10.000 --> 00:23:11.000 So the management of the... 00:23:11.000 --> 00:23:14.000 So one possibility is we have new top-level domains which are managed completely 00:23:14.000 --> 00:23:17.000 differently, whereas we'll find ways of ruling out speculation. 00:23:18.000 --> 00:23:25.000 Another possibility is we actually start creating... replacing the use of the 00:23:25.000 --> 00:23:27.000 domain name system by something else. 00:23:27.001 --> 00:23:29.001 Exciting times, exciting times. 00:23:29.001 --> 00:23:32.000 And we will all end up... 00:23:32.000 --> 00:23:34.001 It's very important that we all end up producing standards. 00:23:35.000 --> 00:23:39.001 Of course, some of the exciting debate is going to be about the way these 00:23:39.001 --> 00:23:44.000 different projects can actually use different pieces of each other. 00:23:44.001 --> 00:23:49.000 It's hard to do that, because it involves constraining the way you work to be 00:23:49.000 --> 00:23:53.001 able to be compatible with the way somebody else works, which sucks, because it's 00:23:53.001 --> 00:23:57.000 so much easier to just go straight ahead and design your own system without 00:23:57.000 --> 00:24:00.000 looking at either side, and then in fact you always have to catch it. 00:24:01.000 --> 00:24:03.000 You have to hit a happy medium. 00:24:03.001 --> 00:24:05.001 Other things that... 00:24:05.001 --> 00:24:10.001 Oh, by the way, you have things I should mention happening in W3C, the social web 00:24:10.001 --> 00:24:14.000 working group you should check out, which is very much in this space. 00:24:14.000 --> 00:24:17.001 It's about... very much about decentralized power. 00:24:18.000 --> 00:24:20.000 Web authentication is a biggie. 00:24:20.000 --> 00:24:21.000 All these systems need authentication. 00:24:21.001 --> 00:24:25.000 There's a web authentication working group which has taken over the FIDO work. 00:24:25.001 --> 00:24:26.001 There's also a working group... 00:24:26.001 --> 00:24:33.000 There's a community group doing hardware tokens which you could check out. 00:24:33.001 --> 00:24:38.000 So those things are close by, you know, in places where to find other 00:24:38.000 --> 00:24:39.001 like-minded thinking people. 00:24:41.001 --> 00:24:48.001 I think that also one of the things I will see is emerging of the 00:24:48.001 --> 00:24:55.000 world of synchronization, of sync and synchronization systems, things which sync 00:24:55.000 --> 00:25:00.000 your phone and your laptop and bits of the cloud should start to be 00:25:00.000 --> 00:25:02.001 indistinguishable from the web. 00:25:03.001 --> 00:25:05.001 Yeah, the web itself also... 00:25:05.001 --> 00:25:08.000 A lot of the websites out there are actually behind them. 00:25:08.001 --> 00:25:09.000 There's a source code system. 00:25:09.001 --> 00:25:12.000 Like, you can see the website and then you can also follow a link 00:25:12.000 --> 00:25:13.001 to the GitHub repository. 00:25:14.001 --> 00:25:17.001 Where is the Git source code control of all 00:25:17.001 --> 00:25:19.000 the different versions of that website? 00:25:19.001 --> 00:25:25.000 Maybe we should start introducing to HTTP the links so that actually we start 00:25:25.000 --> 00:25:27.000 surfacing the whole... 00:25:27.000 --> 00:25:34.000 We basically surface Git or a redesign of Git that's easier 00:25:34.000 --> 00:25:39.001 to use, which we surface [...] code control functionality so that we capture all 00:25:39.001 --> 00:25:42.001 the different versions and they're all visible in a standard way. 00:25:43.000 --> 00:25:43.001 Wouldn't that be cool? 00:25:44.000 --> 00:25:50.000 So when we do that, then we end up having common languages for 00:25:50.000 --> 00:25:51.001 talking about the differences between two versions. 00:25:52.000 --> 00:25:54.000 And when you can have common languages between talking about the differences 00:25:54.000 --> 00:25:58.000 between two versions, there are really cool things you can do by passing all the 00:25:58.000 --> 00:26:03.000 differences in a document that you have on your laptop between around the other 00:26:03.000 --> 00:26:06.000 people in the boardroom as you're editing something so that you don't 00:26:06.000 --> 00:26:07.001 have to even rely on... 00:26:07.001 --> 00:26:10.001 You can use the same technology, but actually the systems could talk to each 00:26:10.001 --> 00:26:12.001 other over Bluetooth instead of... 00:26:12.001 --> 00:26:13.001 Because they can. 00:26:14.000 --> 00:26:20.000 Lots of really exciting possibilities and they connect together in ways which 00:26:20.000 --> 00:26:23.001 would mean to try to elaborate all the connections would take much too long. 00:26:23.001 --> 00:26:28.000 I just wanted to throw in some of the ways in which I'm A, frustrated on behalf 00:26:28.000 --> 00:26:32.001 of everybody who's using the web at the moment and stuck... 00:26:33.001 --> 00:26:40.000 The silos be very excited that we can re-decentralise the web 00:26:40.001 --> 00:26:45.000 and very excited that we've got this incredible group of people here and very 00:26:45.000 --> 00:26:50.000 thankful to Brewster and company at the archive 00:26:50.000 --> 00:26:52.000 for hosting it or putting it all together. 00:26:52.000 --> 00:26:53.000 Thank you very much. 00:26:53.000 --> 00:26:53.001 Thank you. 00:26:57.001 --> 00:26:58.000 APPLAUSE 00:27:01.000 --> 00:27:02.001 Shall we take some questions? 00:27:03.000 --> 00:27:06.000 If you have a question for Sir Tim, please come to this mic. 00:27:08.000 --> 00:27:11.001 I remind everyone to make it a question, not a statement. 00:27:13.000 --> 00:27:14.001 And ask one question at a time. 00:27:15.000 --> 00:27:16.000 So we can have more people. 00:27:17.000 --> 00:27:17.001 Shall I go up at the end? 00:27:18.001 --> 00:27:20.000 I guess first, happy birthday. 00:27:21.001 --> 00:27:22.000 Thank you. 00:27:25.001 --> 00:27:26.000 APPLAUSE 00:27:30.001 --> 00:27:37.001 And I guess I'm heartened by the fact that so far lots of the talks 00:27:37.001 --> 00:27:41.001 have been about the values that we want and how we can sort of define what we 00:27:41.001 --> 00:27:44.000 want ahead of time before we go out and try to build it. 00:27:44.001 --> 00:27:48.001 But I also wonder if you could comment just on maybe if you could just stay the 00:27:48.001 --> 00:27:54.000 top three maybe promising projects that achieve a lot of the 00:27:54.000 --> 00:27:55.001 things we've already talked about. 00:27:56.000 --> 00:27:58.001 Like name data networking, IPFS things like that. 00:27:59.000 --> 00:28:00.000 Oh, you know what? 00:28:00.001 --> 00:28:03.001 People ask me for my top three websites and I always refuse 00:28:04.001 --> 00:28:06.000 because they're all very wonderful. 00:28:07.000 --> 00:28:09.001 So I asked the top three... 00:28:09.001 --> 00:28:11.001 There are... 00:28:11.001 --> 00:28:14.000 I would... 00:28:14.000 --> 00:28:16.001 Sorry, I'm not going to pick a top three because 00:28:16.001 --> 00:28:18.000 they are wonderful in different ways. 00:28:18.000 --> 00:28:22.001 It's not that they're all trying to achieve one particular thing and it's a 00:28:22.001 --> 00:28:24.000 question of who's closest to the moon? 00:28:25.000 --> 00:28:28.001 They've got one person as with the moonshot, we've got some people working on 00:28:28.001 --> 00:28:31.001 fuel, we've got some people working on space suits, we've got some people working 00:28:31.001 --> 00:28:34.001 on physics and navigation. 00:28:35.001 --> 00:28:36.000 And they all... 00:28:36.000 --> 00:28:37.001 So the interesting thing is how they all fit together. 00:28:38.000 --> 00:28:41.000 You can't put them in order, just like you can't put people in order. 00:28:43.001 --> 00:28:44.000 APPLAUSE 00:28:47.001 --> 00:28:48.000 Hi, Tim. 00:28:49.001 --> 00:28:53.001 So I got involved a few years ago in trying to keep a bunch of my friends from 00:28:53.001 --> 00:28:58.001 joining a walled garden, but they were non-technical people and it was ultimately 00:28:58.001 --> 00:29:02.001 impossible because the design of that garden made it very easy 00:29:02.001 --> 00:29:04.001 for them to understand what to do. 00:29:05.000 --> 00:29:10.001 So how do we get better design for our better for the world versions of these 00:29:10.001 --> 00:29:12.000 things because that's really the problem? 00:29:14.001 --> 00:29:17.000 It's a really important problem. 00:29:17.001 --> 00:29:18.001 Yes, I agree. 00:29:19.000 --> 00:29:20.000 One of the... 00:29:20.001 --> 00:29:25.001 So when we're using a decentralized system, if you try to log on to a solid 00:29:25.001 --> 00:29:32.001 server, then what you're doing when you first use any of these 00:29:32.001 --> 00:29:37.000 decentralized apps for the first time is you're getting yourself an ID which will 00:29:37.000 --> 00:29:42.000 allow you to use any decentralized app and also to allow them to store data on 00:29:42.000 --> 00:29:45.000 any of the servers you choose, even if you 00:29:45.000 --> 00:29:46.001 buy a freedom box and plug it in at home. 00:29:47.000 --> 00:29:52.001 You can store stuff on there, and so there are lots of possibilities. 00:29:53.000 --> 00:29:58.001 And so how to make the sign-up experience just like you signed up for the 00:29:58.001 --> 00:30:04.000 simplest social networking chat is a really important goal. 00:30:04.000 --> 00:30:08.001 And in a way, of all these experience that they have, the sign-up experience is 00:30:08.001 --> 00:30:11.000 kind of a crucial one to the onboarding. 00:30:11.001 --> 00:30:12.000 So I agree. 00:30:15.000 --> 00:30:15.001 [...] 00:30:15.001 --> 00:30:20.000 So I hope that one of the things that we'll do is we'll end up with really great 00:30:20.000 --> 00:30:24.000 designers pitching in to look at the... 00:30:24.001 --> 00:30:30.000 Just to rewind, just to rant, start my PGP rant. 00:30:30.000 --> 00:30:31.001 PGP is a great system. 00:30:32.001 --> 00:30:33.001 Okay, it's a great... 00:30:33.001 --> 00:30:36.000 The software UI is really horrible. 00:30:36.000 --> 00:30:39.001 It's been worked on by hardly anybody with hardly any funds. 00:30:41.000 --> 00:30:46.001 But, for example, if you were to take the GPG code 00:30:49.000 --> 00:30:55.001 and take the drop-down which says, your key isn't known, you should launch the 00:30:55.001 --> 00:31:00.001 Key Manager and go download it and search for it, and then store it and then save 00:31:00.001 --> 00:31:04.000 it, and then maybe this message will go away and change it to a button which 00:31:04.000 --> 00:31:05.001 just does all that, okay? 00:31:07.000 --> 00:31:11.001 Sometimes the things are very obvious and there are places, systems where for 00:31:11.001 --> 00:31:16.000 years people haven't put the effort into just smoothing it out. 00:31:16.001 --> 00:31:19.000 And sometimes it's not obvious how to... 00:31:19.001 --> 00:31:24.001 And sometimes, yeah, the people, of course, the huge holds of people who work in 00:31:24.001 --> 00:31:29.000 the walled gardens, as gardeners, they are paid well. 00:31:30.000 --> 00:31:37.000 But it always turns out, in the long term, that you can make the 00:31:37.000 --> 00:31:41.001 walled garden very, very sweet, but the jungle outside is always more appealing 00:31:41.001 --> 00:31:45.001 in the long term because of the diversity which it produces. 00:31:46.000 --> 00:31:50.000 So one of the cool things is, in our decentralized world, we will have lots and 00:31:50.000 --> 00:31:54.000 lots of competing apps, and if we have good standards, they will store data in 00:31:54.000 --> 00:31:56.001 compatible forms so you could switch from one to another. 00:31:57.001 --> 00:32:01.001 So keep preserving your entire social graph and your contacts and your 00:32:01.001 --> 00:32:04.000 life and your fitness and all that. 00:32:04.000 --> 00:32:08.001 You will be able to switch, choose which fitness apps and life and contact 00:32:08.001 --> 00:32:12.000 managing apps you can use because we're using standards. 00:32:12.001 --> 00:32:19.001 And so that, in a way, will allow us to... 00:32:19.001 --> 00:32:22.000 Will allow us to... 00:32:22.001 --> 00:32:25.000 Now people move from one to another. 00:32:25.001 --> 00:32:29.000 Okay, next. Thanks. 00:32:30.000 --> 00:32:33.001 Hi. First I want to say thanks for your vague but interesting idea. 00:32:33.001 --> 00:32:35.001 I've got maybe a vague question of my 00:32:35.001 --> 00:32:38.000 own, actually related to what you were just saying. 00:32:39.001 --> 00:32:41.000 So the web as it is now... 00:32:41.000 --> 00:32:41.001 [...] 00:32:43.001 --> 00:32:48.001 The web as it is now is kind of designed for documents, and we still talk about 00:32:48.001 --> 00:32:54.001 web documents, but so much of applications and things like that is really data 00:32:54.001 --> 00:32:56.001 being delivered over the web. 00:32:58.000 --> 00:33:00.001 But there's no interoperable... 00:33:00.001 --> 00:33:03.001 Facebook has its standard for their data. 00:33:04.001 --> 00:33:07.001 Yelp, all these companies have different ways of representing their data. 00:33:08.001 --> 00:33:14.000 Are you aware of any standards that kind of hope to kind of unify this kind of 00:33:14.000 --> 00:33:19.000 data layer, and how would you go about kind of promoting that, 00:33:19.000 --> 00:33:20.001 given these ideas are really hard? 00:33:21.000 --> 00:33:23.001 There's a million different ideas for doing this. 00:33:23.001 --> 00:33:25.001 How do you standardize people on one? 00:33:26.000 --> 00:33:28.001 Well, the Solid Project is exactly that. 00:33:28.001 --> 00:33:31.000 So github.com slash solid. 00:33:32.000 --> 00:33:34.000 It's early days. It's at the developer getting... 00:33:34.000 --> 00:33:39.001 So how would I promote it? I would say if you're a developer, you do things like 00:33:39.001 --> 00:33:45.000 think about using Turtle and drinking the Link Data 00:33:45.000 --> 00:33:46.001 Kool-Aid, because it's cool. 00:33:48.001 --> 00:33:54.000 And then look at the solid stuff, because the solid... 00:33:54.000 --> 00:33:55.001 the whole idea of that is... 00:33:55.001 --> 00:33:57.000 I suppose there's two levels. 00:33:57.001 --> 00:33:59.001 One is the goal... 00:34:00.001 --> 00:34:02.000 Now, I'm talking particularly about this particular project. 00:34:02.001 --> 00:34:09.000 The goal is that the data stores have generic APIs. 00:34:09.001 --> 00:34:11.000 So that is one level of standardization. 00:34:11.001 --> 00:34:15.000 You can tweak files. You can add data. 00:34:15.000 --> 00:34:18.000 You can just add facts to a file 00:34:18.000 --> 00:34:22.001 by... by poking them. 00:34:23.001 --> 00:34:29.000 So at that level that you're reading and writing files and you're also editing 00:34:29.000 --> 00:34:31.001 files, you can treat any file as a database basically. 00:34:32.001 --> 00:34:36.001 That level, the server just implements a certain functionality no more. 00:34:37.000 --> 00:34:38.001 So that we need a certain standardization. 00:34:39.000 --> 00:34:43.001 So to be a solid server, we've got one in Go and we've got one in Node. 00:34:44.001 --> 00:34:47.000 And to be a solid server, you just need that level of standardization. 00:34:47.000 --> 00:34:51.000 But to make it all work, the apps can write stuff 00:34:51.000 --> 00:34:52.001 into the store, but they can't just write anything. 00:34:53.000 --> 00:34:56.000 So then we have to standardize how you store your contacts, how you store your 00:34:56.000 --> 00:35:00.001 friends, how you store your... how you build the interactions, because these 00:35:00.001 --> 00:35:02.001 aren't just pieces of data. 00:35:02.001 --> 00:35:05.001 They're invitations and acceptances and things like that. 00:35:06.000 --> 00:35:08.000 So we're working through that to join us. 00:35:09.000 --> 00:35:10.001 How we encourage... 00:35:10.001 --> 00:35:12.001 I hereby encourage you. 00:35:12.001 --> 00:35:14.000 How do you encourage like this? 00:35:14.001 --> 00:35:15.001 This is the best I can do. 00:35:16.000 --> 00:35:16.001 Thank you. 00:35:18.001 --> 00:35:18.001 Hello. 00:35:18.001 --> 00:35:22.000 Thank you so much for attending and for your work. 00:35:23.000 --> 00:35:28.001 I'm wondering how you feel about global access to the internet and how keeping 00:35:28.001 --> 00:35:30.001 this sort of jungle open allows... 00:35:31.001 --> 00:35:34.001 You know, like Africa is considered a mobile-only market at this 00:35:34.001 --> 00:35:36.001 point rather than mobile first. 00:35:37.000 --> 00:35:41.001 And just the way traditionally disadvantaged populations in the world 00:35:41.001 --> 00:35:43.000 are starting to access the internet. 00:35:44.000 --> 00:35:49.000 I'm wondering if you think like keeping this unsilened internet allows people in 00:35:49.000 --> 00:35:52.001 those areas to have the sort of access that we all had in the 90s. 00:35:53.000 --> 00:35:57.000 And ultimately, you know, everyone in this room were really privileged to be 00:35:57.000 --> 00:35:59.000 embedded in tech to this degree. 00:35:59.001 --> 00:36:03.000 So just I'm wondering if like the projects at MIT and stuff like that, if they 00:36:03.000 --> 00:36:07.000 have any focus on, you know, that sort of access globally. 00:36:08.000 --> 00:36:11.000 Well, check out the Web Foundation, for example, webfoundation.org. 00:36:11.000 --> 00:36:13.001 Yeah, we are right now... we are privileged. 00:36:13.001 --> 00:36:18.001 [...] are... the number of people using the web... 00:36:19.001 --> 00:36:23.001 When we started the Web Foundation, you know, we were shocked that it was 10 00:36:23.001 --> 00:36:25.000 % and 20% of the world. 00:36:25.001 --> 00:36:29.000 And at that point, for the first time, you know, we couldn't have talked about it 00:36:29.000 --> 00:36:32.001 as a human right when it was . 001% of the people in the room. 00:36:33.000 --> 00:36:34.001 But now it's getting to this... 00:36:34.001 --> 00:36:36.001 So we started the Web Foundation because, well, 00:36:37.000 --> 00:36:39.000 what about all those people who don't have access? 00:36:39.000 --> 00:36:42.001 Well, now it's come up to... it's about 41%, roughly. 00:36:43.000 --> 00:36:46.000 So that means that something very cool is going to happen. 00:36:46.001 --> 00:36:48.000 Something very historic is going to happen. 00:36:48.001 --> 00:36:52.000 It's going to... it's going to go from being a soon next year or 00:36:52.000 --> 00:36:53.001 two, it's going to cost 50%. 00:36:53.001 --> 00:36:57.001 And so suddenly what was a minority thing will become a majority thing. 00:36:57.001 --> 00:37:00.000 So yes, obviously, a huge amount of effort. 00:37:00.000 --> 00:37:04.001 As we do that, as we go through this, a lot of effort needs to... 00:37:04.001 --> 00:37:07.001 So they look at the Alliance for the Affordable Internet and things like that, 00:37:08.001 --> 00:37:14.000 A4AI.org, trying to just get the costs down of devices, get the cost down of 00:37:14.000 --> 00:37:18.001 connectivity, making sure the wanton needs more taxes, on monopolies across the 00:37:18.001 --> 00:37:20.000 developing world. 00:37:21.001 --> 00:37:23.001 So that's really, really important as we go through this. 00:37:24.000 --> 00:37:29.001 But then, of course, as it suddenly becomes a majority, there's a different... in 00:37:29.001 --> 00:37:32.000 a way, the issue, perhaps, is something different. 00:37:32.000 --> 00:37:36.000 And then now, oh, the majority of people in the world have the Web, 00:37:36.000 --> 00:37:38.000 so your business or your government doesn't 00:37:38.000 --> 00:37:39.001 have to really bother with the people who don't. 00:37:40.000 --> 00:37:46.000 So the government, for now, the digital divide has become... the gap has become 00:37:46.000 --> 00:37:49.001 bigger because suddenly a lot of people made the assumption that you're online. 00:37:50.000 --> 00:37:56.000 And so for the last 20%, the last 10%, we need to work a lot of mitigation for 00:37:56.000 --> 00:37:58.001 how to help people live who are not online and so on. 00:37:58.001 --> 00:38:04.001 And so, okay, what we'll do is I'll close the 00:38:04.001 --> 00:38:07.000 queue to questions and just take these three questions. 00:38:08.000 --> 00:38:09.001 Thank you very much. 00:38:10.001 --> 00:38:12.001 I said I could... I'll take these... 00:38:12.001 --> 00:38:12.001 Oh, one more. 00:38:13.000 --> 00:38:14.001 I said I'd close the queue to 00:38:14.001 --> 00:38:16.001 questions, but I take the ones that are on the queue. 00:38:16.001 --> 00:38:19.001 Oh, so I think the Web platform... 00:38:19.001 --> 00:38:20.000 My name's Max. 00:38:20.000 --> 00:38:23.001 The Web platform has evolved an amazing amount in the last 10 years. 00:38:24.001 --> 00:38:25.001 JavaScript is a lot faster. 00:38:25.001 --> 00:38:29.001 But I think that a lot of the fundamental innovations have kind of been 00:38:29.001 --> 00:38:32.000 accidental or driven by big interests. 00:38:33.000 --> 00:38:36.000 So, for example, we got WebGL because people wanted to make games on the browser, 00:38:36.001 --> 00:38:40.001 and then now we get, as a side effect, typed arrays, and we can work with binary 00:38:40.001 --> 00:38:42.001 data, which opens up a whole new section of Web apps. 00:38:43.000 --> 00:38:47.000 And then people wanted to make video calls on the browser, so we got WebRTC, and 00:38:47.000 --> 00:38:50.000 as a side effect, we can build these distributed file systems in the browser. 00:38:50.001 --> 00:38:54.001 Do you think that we need, as a group, one of these big interests to push through 00:38:54.001 --> 00:38:59.000 the standards process that secretly gives us these side effects that we need? 00:38:59.000 --> 00:39:03.000 Or do you think, as a group, we don't have the use case yet that can get the 00:39:03.000 --> 00:39:07.000 standards implemented because we need a kind of big semi-truck 00:39:07.000 --> 00:39:08.001 to ride in on? Do you think that's accurate? 00:39:09.001 --> 00:39:11.000 I don't think we need something weird. 00:39:11.001 --> 00:39:15.000 One, I use this stuff for my daily life. 00:39:15.001 --> 00:39:22.000 I use it for my... For example, if you invite me to something... to a seminar, it 00:39:22.000 --> 00:39:24.001 goes into a tracker, which is all based on solid. 00:39:25.000 --> 00:39:29.001 And so that's been... I've been using that basically for years and keeping it 00:39:29.001 --> 00:39:31.001 tracking the solid standard. 00:39:32.000 --> 00:39:37.000 And my feeling is that if you want the sort of killer app, it may... 00:39:37.000 --> 00:39:41.000 One of the things... So one possibility is actually the early adopter community 00:39:41.000 --> 00:39:47.001 is developers because they can install stuff, and also they will understand 00:39:47.001 --> 00:39:54.001 what's going on and maybe just... We have used IRC for ages 00:39:54.001 --> 00:39:57.001 because it's been kind of... had a nice decentralized, uncontrolled feel. 00:39:58.000 --> 00:40:01.001 But I think there's a Slack channel here. 00:40:02.000 --> 00:40:04.000 No, Slack is another side of... 00:40:04.001 --> 00:40:06.001 You're recommended to tweet. 00:40:07.000 --> 00:40:11.000 Twitter is tricky because it's a big... It's centralized for a reason. 00:40:11.001 --> 00:40:13.000 Slack isn't centralized for a reason. 00:40:13.000 --> 00:40:16.001 It's centralized just in the lots and lots of versions of Gitcher and stuff. 00:40:16.001 --> 00:40:20.001 So one possibility is we just get together afterwards tomorrow and we make a 00:40:20.001 --> 00:40:27.001 solid-based chat and basically whatever groups need to 00:40:27.001 --> 00:40:31.001 organize and share meeting materials and stuff like that. Organize... A few 00:40:31.001 --> 00:40:33.001 basic... 00:40:33.001 --> 00:40:37.000 So it could be that it's not rocket science apps, but just day-to-day life really 00:40:37.000 --> 00:40:38.001 useful apps could do it. 00:40:39.000 --> 00:40:39.001 Cool. Thanks. 00:40:42.000 --> 00:40:42.001 Hey, my name is Juan. 00:40:45.001 --> 00:40:48.000 My question is going to be about kind of the frontier. 00:40:48.001 --> 00:40:52.001 So we're here talking about decentralization and location and moving the web 00:40:52.001 --> 00:40:54.001 outside of some particular point and 00:40:54.001 --> 00:40:56.000 adding content addressing and all that kind of stuff. 00:40:56.000 --> 00:40:59.000 But there's a lot of other things going on in the network right now. 00:40:59.001 --> 00:41:03.001 So things like Bitcoin reminded us that we can have payment systems that are 00:41:03.001 --> 00:41:07.001 decentralized. We are getting overturned in terms of the organizations 00:41:07.001 --> 00:41:09.001 or authorities that run these systems. 00:41:11.001 --> 00:41:16.000 Web 1.0 was about connecting all these devices to be able to exchange documents 00:41:16.000 --> 00:41:17.001 and be able to have information sharing. 00:41:18.000 --> 00:41:21.001 Web 2.0 was about making everything dynamic and being able to have a read-write 00:41:21.001 --> 00:41:23.001 connection with the websites. 00:41:24.001 --> 00:41:26.000 What do you think is Web 3.0? 00:41:26.000 --> 00:41:30.000 There's things like the blockchain that are giving us payment systems and 00:41:30.000 --> 00:41:32.001 application processing systems that are decentralized. 00:41:33.000 --> 00:41:36.001 There's things like the file systems that we're going to talk about in this 00:41:36.001 --> 00:41:39.000 conference that gives us... that lifts the web from location. 00:41:40.000 --> 00:41:43.000 There's a lot of disenfranchised people around the planet, people in developing 00:41:43.000 --> 00:41:46.001 world that can't actually connect reliably to the rest of the internet, cannot 00:41:46.001 --> 00:41:48.001 maintain a proper connection. 00:41:49.001 --> 00:41:51.000 What do you think is the frontier? 00:41:52.000 --> 00:41:54.000 We'd love to hear more about various ideas. 00:41:55.000 --> 00:42:00.001 So for that piece, there are solutions coming along and low 00:42:00.001 --> 00:42:02.000 -earth with satellites and things like that. 00:42:02.001 --> 00:42:03.000 That's a bit of a distraction. 00:42:03.000 --> 00:42:07.001 Let's go back to the essential thing about the web. 00:42:10.000 --> 00:42:10.001 [...] 00:42:10.001 --> 00:42:12.000 Yes, the web is a web of documents. 00:42:13.001 --> 00:42:19.000 When you download an app or web app, when it comes... and you 00:42:19.000 --> 00:42:21.001 run it, you're running a bunch of JavaScript. 00:42:22.000 --> 00:42:24.001 Very often, you're actually exploring a piece of data. 00:42:25.000 --> 00:42:27.000 So in fact, behind that is the web of data. 00:42:27.001 --> 00:42:31.001 So one of the directions is exposing the web of data to users, if you like, 00:42:31.001 --> 00:42:35.001 allowing them to realize that when they're looking at a bank statement presented 00:42:35.001 --> 00:42:39.001 by their bank, they could actually turn on a bunch of really cool, more powerful 00:42:39.001 --> 00:42:43.000 tools to look at that bank statement and to do their taxes and so on, and they 00:42:43.000 --> 00:42:45.000 could buy and pay good money for apps. 00:42:45.001 --> 00:42:51.001 So exposing the web of data is, in a way, to a certain extent, is what 00:42:51.001 --> 00:42:53.000 part of the answer to the app question. 00:42:53.000 --> 00:42:57.000 The reason why people have apps is because they need powerful ways of looking at 00:42:57.000 --> 00:42:59.001 the data, and in fact... 00:42:59.001 --> 00:43:03.001 So maybe your will be a web which is two pieces. 00:43:04.000 --> 00:43:08.001 It's just a plain vanilla document that you will just look at with a normal 00:43:08.001 --> 00:43:14.001 browser and the space of data which you explore using powerful tools, but 00:43:14.001 --> 00:43:20.000 tools which will be hopefully archivable and 00:43:20.000 --> 00:43:25.001 which will be looking at standard data, which you 00:43:25.001 --> 00:43:27.000 will still be able to read in 20 years' time. 00:43:28.000 --> 00:43:28.001 [...] 00:43:29.000 --> 00:43:30.000 So the post... 00:43:30.000 --> 00:43:34.001 I wasn't going to take a good question, but this fits so well with your answer. 00:43:36.000 --> 00:43:37.000 Linked data. 00:43:37.000 --> 00:43:40.000 I've watched the libraries and archives world put a lot of effort into adopting, 00:43:40.000 --> 00:43:43.001 getting their heads around linked data and then getting momentum up behind 00:43:43.001 --> 00:43:49.001 adoption, and now when I go into that space and talk about decentralization, some 00:43:49.001 --> 00:43:55.001 confusion comes up around identifiers, and I wonder, in the rules of linked data, 00:43:56.000 --> 00:44:01.000 one of the rules is that your identifiers are HTTP URIs, where HTTP was designed 00:44:01.000 --> 00:44:03.001 for the web of documents, right? 00:44:04.000 --> 00:44:05.000 Hypertext, right? 00:44:05.001 --> 00:44:11.001 So do you feel like HTTP still has a role in the web of 00:44:11.001 --> 00:44:13.000 data or... 00:44:13.000 --> 00:44:14.001 No, no, no, no, no. 00:44:15.001 --> 00:44:21.001 The HTTP was not designed for a web of documents, in the sense of word documents. 00:44:22.000 --> 00:44:26.001 HTTP was designed to be a set of identifiers for anything, and what semantic web 00:44:26.001 --> 00:44:28.001 says is you can use... 00:44:28.001 --> 00:44:34.000 You can make a URI to talk about this church, this organization yourself. 00:44:35.000 --> 00:44:36.001 So if you... 00:44:36.001 --> 00:44:41.001 In fact, so quick web architecture 101. 00:44:41.001 --> 00:44:44.001 The hash, the web architecture is the hash. 00:44:45.001 --> 00:44:50.001 The web design is we will take the identifiers of documents and we will put a 00:44:50.001 --> 00:44:54.001 hash, and then after the hash, we will put something else, which depending on 00:44:54.001 --> 00:44:57.001 what system you're in, identifies something within the document. 00:44:58.000 --> 00:45:01.001 So this can be a document about a church, and then this church itself 00:45:01.001 --> 00:45:03.000 is the church itself. 00:45:04.000 --> 00:45:11.000 So RDF uses HTTP differently to the way HTML uses it, and it 00:45:11.000 --> 00:45:12.001 [...] 00:45:12.001 --> 00:45:19.000 So you can use these identifiers just as identifiers. 00:45:19.000 --> 00:45:20.001 You can use them like hashes if you like. 00:45:20.001 --> 00:45:22.001 Oftentimes you can look, you can cache them, and you can hash them, and 00:45:22.001 --> 00:45:24.000 you can... 00:45:24.000 --> 00:45:28.001 And in fact, because they start with HTTP, you can look them up as well. 00:45:28.001 --> 00:45:31.000 But that's something that you always need. 00:45:32.000 --> 00:45:34.000 We can change... 00:45:34.000 --> 00:45:35.001 Yeah, sometimes you need different... 00:45:35.001 --> 00:45:39.000 Sometimes you need to have URNs, but most of the time, almost 00:45:39.000 --> 00:45:40.001 always you need to have something... 00:45:40.001 --> 00:45:43.000 It's valuable to look these things up. 00:45:43.000 --> 00:45:48.000 So I think what I'd like to be able to do is to stick with using HTTP 00:45:50.001 --> 00:45:55.001 for the namespace, and then morph, and then we're constantly changing. 00:45:56.000 --> 00:45:59.000 HTTP is constantly changing. HTTP2 is a complete redesign, just came out. 00:45:59.001 --> 00:46:03.001 Okay, we're constantly adding headers to it in a solid project, in other 00:46:03.001 --> 00:46:06.001 projects, because you can throw in your other headers. 00:46:06.001 --> 00:46:13.000 So if you want to use it in a particular way for semantic web identifiers or 00:46:13.000 --> 00:46:17.001 things which are distributed between different desktops, or for example, or in a 00:46:17.001 --> 00:46:20.001 very peer-to-peer way, you can do that. 00:46:22.000 --> 00:46:24.000 So anyway, we can talk about... Let's talk about this at the break. 00:46:24.001 --> 00:46:25.000 Thank you. 00:46:31.001 --> 00:46:34.001 Let's give it up for Sir Tim Berners-Lee on his birthday. 00:46:36.001 --> 00:46:37.001 Yay! 00:46:42.000 --> 00:46:43.001 For the history books...